Deletion of the NMDA-NR1 receptor subunit gene in the mouse nucleus accumbens attenuates apomorphine-induced dopamine D1 receptor trafficking and acoustic startle behavior

Abstract

The nucleus accumbens (Acb) contains subpopulations of neurons defined by their receptor content and potential involvement in sensorimotor gating and other behaviors that are dysfunctional in schizophrenia. In Acb neurons, the NMDA NR1 (NR1) subunit is coexpressed not only with the dopamine D1 receptor (D1R), but also with the µ-opioid receptor (µ-OR), which mediates certain behaviors that are adversely impacted by schizophrenia. The NMDA-NR1 subunit has been suggested to play a role in the D1R trafficking and behavioral dysfunctions resulting from systemic administration of apomorphine, a D1R and dopamine D2 receptor agonist that impacts prepulse inhibition to auditory-evoked startle (AS). Together, this evidence suggests that the NMDA receptor may regulate D1R trafficking in Acb neurons, including those expressing µ-OR, in animals exposed to auditory startle and apomorphine. We tested this hypothesis by combining spatial-temporal gene deletion technology, dual labeling immunocytochemistry, and behavioral analysis. Deleting NR1 in Acb neurons prevented the increase in the dendritic density of plasma membrane D1Rs in single D1R and dual (D1R and µ-OR) labeled dendrites in the Acb in response to apomorphine and AS. Deleting NR1 also attenuated the decrease in AS induced by apomorphine. In the absence of apomorphine and startle, deletion of Acb NR1 diminished social interaction, without affecting novel object recognition, or open field activity. These results suggest that NR1 expression in the Acb is essential for apomorphine-induced D1R surface trafficking, as well as auditory startle and social behaviors that are impaired in multiple psychiatric disorders.

Figure 1

NR1 deletion in the Acb attenuates the normal apomorphine-induced reduction in the peak amplitude of auditory-evoked startle. Apomorphine results in a reduction in AS (A) but not %PPI (B) in fNR1 mice receiving Acb rAAV-GFP. There is no reduction in AS (C) or %PPI (D) in fNR1 mice receiving rAAV-Cre in the Acb. *p<0.05, saline versus apomorphine in Acb rAAV-GFP injected mice by Fisher’s PLSD.

Figure 2

Immunoperoxidase labeling of the GFP reporter protein in the Acb following local unilateral microinjection of vector into either the core (A) or shell (B) subregions. Areas bounded by the trapezoids represent regions where tissue was sampled by EM in adjacent sections processed for dual D1R and μ-OR immunolabeling. (C-F). NR1 immunolabeling in the Acb core (C, E) and shell (D, F) of rAAV-GFP and rAAV-Cre injected mice, respectively. Bar=500 μm ac: anterior commissure, CPu: caudate-putamen, D: dorsal, lv: lateral ventricle; M: medial.

Figure 3

Immunogold D1R and immunoperoxidase μ-OR labeling in the Acb core of mice receiving rAAV-GFP and treated with saline (n=3) or apomorphine (Apo, n=3) and tested for AS. In electron micrographs from the saline-treatment group (A-B) the dendritic D1R immunogold particles have a largely cytoplasmic distribution (arrows). In those receiving apomorphine (D-E) D1R immunogold particles show a plasmalemmal location (circles) in single (D1-den) and dual labeled D1R/μ-OR-den dendrites (D1/μ-OR-den). In both the saline and apomorphine treated mice, μ-OR containing dendritic spines (μOR-sp) receive asymmetric excitatory-type synapses (arrow heads) from unlabeled terminals (U-te), but the representative dendritic spines appear larger in the mice receiving apomorphine (F) compared with those receiving saline (C). In G, the bar graph shows a significant increase in surface labeling in single D1R and dual labeled dendritic shafts, as well as D1R labeled spines. In H, there is a significant increase in the cross-sectional area of μ-OR labeled spines in the Acb core of apomorphine treated animals. Scale bars = 500 nm. Data shown are the mean±SEM for each group. *p<0.005, Apo relative to saline,Tukey’s HSD.

Figure 4

Single and dual labeled dendrites and spines in the Acb core of mice receiving local injections of rAAV-GFP (n=3) or rAAV-Cre (n=3) and treated with apomorphine (Apo) and AS. Electron microscopic images (A-C) show plasmalemmal (circles) and cytoplasmic (arrows) immunogold labeling in D1R (D1-den) and dual labeled dendrites (D1/μ-OR-den), as well as a single D1R labeled spine (D1-sp). Bar graphs in D and E quantitatively confirm that there is no difference in the density of surface D1R or spine size in Acb neurons of apomorphine treated mice after local NR1 deletion. However, there is a significant decrease in the size of dual labeled dendritic profiles in the Acb of NR1 KO mice given apomorphine. Scale bars = 500 nm. Data shown are the mean±SEM for each group. *p<0.05 in Cre Apo relative to GFP Apo groups by Tukey’s HSD.

Figure 5

Acb NR1 deletion impairs sociability. Mice injected with rAAV-Cre bilaterally in the Acb (Acb Cre) differ in sociability when compared to those receiving Acb rAAV-GFP (Acb GFP) but not rAAV-Cre outside the Acb (i.e. placement control [PC] Cre). As shown in A, mice receiving Cre in the Acb (n=11) show significantly less contact time with a novel mouse relative to animals receiving Acb GFP (n=5), however Acb GFP mice do not differ from the PC Cre group (n=3). Acb Cre-injected mice also have longer latencies to initiate contact with the novel mouse relative to Acb GFP treated animals (B), although this fails to achieve statistical significance. *p<0.05 Acb Cre relative to Acb GFP by Fisher’s PLSD.

Figure 6

Contact time with familiar and novel objects during a two-choice novel object recognition test in Acb Cre (n=11), Acb GFP (n=5), and PC Cre mice (n=3). Contact time with objects one (Obj 1) and two (Obj 2) during the familiarization step are presented at the left and subsequent contact time with one of the latter objects (Familiar) and a new object (Novel) during the test phase are presented at the right. None of the groups differed in the amount of time spent with each of the two objects during the familiarization stage. Mice receiving Acb Cre or Acb GFP spend significantly more time engaged with the novel object during the test phase. The PC Cre mice spend similar amounts of time with both objects. *p<0.05 Acb Cre relative to Acb GFP in contact with a novel versus familiar object in Acb GFP and Acb Cre mice by Fisher’s PLSD.

Figure 7

Open field behavior in Acb NR1 KO and control mice. rAAV-Cre injected (Acb Cre n=11) and control (Acb GFP n=5, and PC Cre n=4) mice display no significant differences in locomotor activity as measured by either distance traveled (A) or activity time (B) in an automated open field. Data shown are the mean±SEM for each group.